CPU, APU & Chipsets News - Page 38

AMD's next-gen Ryzen 7000 series "Zen 4" CPUs are expected to pack quite a few surprises, but we're now hearing about second-generation V-Cache and that Zen 4 V-Cache CPUs will be 30% faster than standard Zen 4 CPUs. These CPUs should be what will eventually turn into the Ryzen 9 7950X3D and Ryzen 7 7800X3D processors, with second-gen AMD 3D V-Cache.

In a new video, Moore's Law is Dead says that after talking to some of his sources, that Zen 4 is packing second-generation V-Cache, succeeding the first-gen 3D V-Cache that AMD used on its Ryzen 7 5800X3D processor. The new second-gen V-Cache is expected to provide 30% more performance over a regular Zen 4 processor, which would be bloody awesome.

AMD delivered around 10-15% more performance with Zen 3 V-Cache over the regular Zen 3 CPUs -- the Ryzen 7 5800X3D and Ryzen 7 5800X in this case -- while MLID reports that he has internal AMD comparison that the company did between early Zen 4 (A0) chips and final Zen 3 silicon.

Intel has revealed more details on its upcoming 3D Foveros packaging technology, which it will be using to built its next-gen Meteor Lake, Arrow Lake, and Lunar Lake CPUs of the future.

The company discussed more details on its packaging technology during Hot Chips 34, where we're finding out about all the new exciting GPU technology from the likes of AMD, NVIDIA, Chinese GPU maker Birentech, and more. Intel talked about the work it has been putting into the development of its 14th Gen Core "Meteor Lake" CPUs, with its new 3D Foveros packaging technology.

Intel's new Foveros packaging technology is a next-gen inter-die chiplet packaging technology that Intel will be splitting into 3 different ways: The standard design for high-yield, high-volume production, Foveros Omni that mixes and matches tiles in the base die complex, with up to 4x interconnect bump density, and finally there's Foveros Direct that has 16x interconnect density, with lower latency, higher bandwidth, and lower power/die requirements.

NVIDIA announced its new Grace CPU and Grace Superchip CPU earlier this year at GTC 2022, but now the company has unveiled new details on its new Grace CPU, Orin SoC, and NVLink chip interconnects at Hot Chips 34.

The new NVIDIA Grace CPU is the first CPU from NVIDIA, packing 72 Arm v9.0 cores that support SVE2 and multiple virtualization extensions including Nested Virtualization and S-EL2. NVIDIA is fabricating its Grace CPU on TSMC's new 4N process node: an optimized version of TSMC's 5nm process node, made exclusively for NVIDIA... just like its new Hopper H100 GPU.

NVIDIA designed its new Grace CPU to be used in conjunction with its C2C (Chip-To-Chip) interconnect, where NVLINK is used to make the Superchips, removing all bottlenecks that you'd get with regular cross-socket configurations.

Intel needs 7000 workers to construct what it calls the "largest silicon manufacturing location on the planet", which will be built and operational in Ohio, USA in the future.

In a new report from the Associated Press, we find out that Intel needs 7000+ workers for its new fabrication plant in Ohio, which will cost $20 billion to make... but there are other issues that Intel is facing. A labor shortage is hurting the construction industry that is making it hard on the company to hire the people necessary for construction.

On top of that, the company is also facing more competition for workers that could have flow-on effects on the home building industry in Ohio, which is required for the continuously expanding workforce that Intel is trying to attract for its new plant. Not only are workers required, but they will bring families and form new communities... new houses are required, but workers are constructing the $20 billion plant for Intel.

AMD's new Ryzen 7000 series "Zen 4" processors will launch next month, so now that retail versions of the CPUs are hitting reviewers, engineering samples (ES) variants are hitting the black market.

An alleged AMD Ryzen 5 7600X "Zen 4" engineering sample (ES) has turned up on the Chinese black market, with the seller noting the CPU has a base clock of 4.4GHz, much lower than the retail specs. AMD is expected to have a 4.7GHz base clock on the Ryzen 5 7600X processor, while boost clocks are expected to reach 5.3GHz on the retail CPU.

AMD's new Ryzen 5 7600X engineering sample (ES) processor is listed for 9999 RMB, which works out to just under $1500 without shipping. The price isn't solid as the seller requests direct contact with them before the purchase is made final and you get the engineering sample Ryzen 5 7600X chip sent to you. I don't know why you would when it requires a new AM5 socket motherboard... which you can't buy just yet.

Intel has detailed a little more information on its next-gen CPU packaging technology, shifting some of the dust across the rumors, and more at Hot Chips 34.

Intel showed off a new chip layout for its Meteor Lake platform, with a great look at tiles (or chiplets) including the quad-tile layout: CPU Tile, GPU Tile, SoC Tile, and the IOE tile. Intel confirmed that the CPU Tile would be made on the "Intel 4" process node or 7nm EUV process node, while the SoC Tile and IOE Tiles will be made by TSMC on their 6nm process node (N6).

As for the GPU Tile, Intel Meteor Lake GPUs were rumored for TSMC's upcoming 3nm process node, but due to some issues with that, Intel decided to use TSMC's new 5nm process node instead. Industry sources have said that it was the plan all along: for Intel's GPU Tile on their Meteor Lake CPUs to be on the TSMC 5nm process node.

We know that AMD will be unveiling its next-gen Ryzen 7000 series "Zen 4" processors on August 29, but previous rumors had a launch that was just two weeks later... but then we started hearing about a new, delayed September 27 release.

VideoCardz reported that NDAs were sent out to reviewers, but AMD needed reviewers to re-sign the NDA as the launch of the Ryzen 7000 series "Zen 4" processors had been delayed. In a post on the Chiphell forums, we're hearing that one of the reviewers on Chiphell confirmed that they were told about the updated NDA, and Zen 4 launch being postponed.

The post continues, asking if we remember how long it took for the BIOS issues ot be solved after AMD's current-gen Zen 3 launched. AMD does have its new AM5 socket, has added DDR5 memory and PCIe 5.0 technology support on its new X670E + X670 motherboards, ready for the world of Ryzen 7000 series "Zen 4" CPUs.

Alphacool has confirmed that its family of CPU coolers and AIO cooling solutions are ready for AMD's upcoming AM5 socket.

AMD's upcoming AM5 socket will be the new home for their next-gen Ryzen 7000 series "Zen 4" processors, with AM5-based motherboards rolling out with DDR5 memory and PCIe 5.0 technology for next-gen PCIe 5.0-capable SSDs and GPUs. Alphacool is ready for the new Ryzen 7000 series "Zen 4" processors, with the company saying that their coolers are ready "just in time for the launch of the new AMD processors in September 2022".

We know that AMD's new AM5 socket will use the same socket size as AM4, so we should see virtually all AM4 coolers working with new AM5 motherboards, without many (if any) modifications at all. Most CPU cooling companies are confirming that their CPU coolers and AIO cooling solutions would work on the new AM5 socket, with Alphacool joining the AM5-ready ranks of GIGABYTE, Arctic Cooling, MSI, and Noctua.

Intel's new 13th Gen Core "Raptor Lake" CPU is in the headlines once again in flagship form, where we saw the Core i9-13900K using 350W+ of power but now the Core i9-13900KF (note the F, so no integrated GPU) being pushed up to an insane 6.2GHz overclocked with a water chiller.

Extreme Player pushed the next-gen Intel Core i9-13900KF processor up to 6.2GHz using a custom water chiller setup, inside of an ASUS ROG Maximus Z690 Apex motherboard. The overclocker was able to push the P-Cores up to 6.2GHz all-core under CPU-Z and 6.8GHz all-core on Cinebench R23.

As for benchmark scores, the Cinebench R23 benchmark run on the current-gen flagship Core i9-12900K processor scores around 26,939 points in the multi-core benchmark. The overclocked Intel Core i9-13900K processor when pushed to 6.2GHz with the water chiller keeping it cool, was benching an incredible 42,790 points: nearly 16,000 more points than the 12900KF.

Intel's new 13th Gen Core "Raptor Lake" CPUs will be announced on September 27 -- alongside AMD's new Ryzen 7000 series "Zen 4" processors -- but you'd better be preparing a PSU upgrade if you grab the Core i9-13900K flagship Raptor Lake CPU.

In a new post by ProHardver, we find out that the next-gen Intel Core i9-13900K processor can consume up to 350W pretty damn easily, as long as you have the right CPU cooler, the right 700-series motherboard, and the overclocking skills to do so. The post does make a point of the 350W overclocking mode needing "serious cooling solutions".

The new "extreme 350W performance mode" on the Intel 13th Gen Core "Raptor Lake" flagship CPU shouldn't surprise anyone at all, especially if you're up to date on the leaks and rumors. It was just over a month ago now that I was reporting on a "pre-production" Core i9-13900K processor that was "fully reviewed" and using a huge 420W of power.